Which Foods Change Semen Parameters, and Which Only Correlate

Only one food group has been tested the way a medicine would be: nuts. Two randomised trials in healthy young men, one with whole walnuts (Robbins 2012, 117 men, 12 weeks) and one with a mix of walnuts, almonds and hazelnuts (the FERTINUTS trial, Salas-Huetos 2018, 119 men, 14 weeks), both found higher sperm motility, vitality and normal morphology in the nut group than in men who kept their usual diet without nuts. Neither trial measured pregnancy, and neither recruited men with a fertility problem.

Pooled in a 2024 meta-analysis (Cardoso and colleagues), the two trials gave moderate effects on motility, morphology and vitality and no clear effect on sperm concentration. In FERTINUTS, total sperm count rose in the nut group and sperm DNA fragmentation fell, while reactive oxygen species, chromosome anomalies and DNA methylation did not differ. In the walnut trial, sperm aneuploidy fell within the walnut group over 12 weeks but did not differ between the groups at the end.

Two features of these trials matter when reading a diet sheet. The men were students eating a Western-style diet, not fertility-clinic patients, so the result cannot be assumed to transfer to a man with a low count. And the comparison was nuts added to the usual diet against no nuts at all, so the trials cannot say that nuts beat any other food, only that adding them changed the semen analysis over about three months. The 2024 review also noted that observational studies of men eating smaller amounts of nuts, under about one serving a day, found no association with semen parameters at all.

Beyond nuts, two Italian trials of a Mediterranean-style diet in healthy young men are described in the Agarwal 2025 review: one combined diet with physical activity over 16 weeks in 263 men and reported better total and progressive motility than a leaflet-only control; the other compared a Mediterranean with a low-fat diet over six months in 160 men and reported higher count and concentration only in the Mediterranean group. Those trials tested a whole pattern, not a food, and they were not retrieved for this page; we report them as the review does.

Everything in this section comes from studies that asked men what they ate and then looked at their semen analysis. That design can show that men who eat more fish have higher counts; it cannot show that the fish did it, because men who eat more fish differ from other men in ways a questionnaire does not capture.

Fish is the clearest of these associations. In 155 men at a Boston fertility clinic who gave 338 semen samples over 18 months (Afeiche 2014, the EARTH study), adjusted mean total sperm count rose from 102 million in the lowest quarter of fish intake to 168 million in the highest, and the share of normally shaped sperm was 5.9% versus 7.5% in the same two quarters. In 162 men assessed for infertility at Christian Medical College Vellore (Chakraborty 2024), the one Indian dataset found for this page, men whose eating fitted a 'coastal' pattern of fish, seafood, cereals, beans and coconut oil were less likely to have an abnormal semen analysis (adjusted odds ratio 0.56, 95% CI 0.37 to 0.82). Because that was a pattern rather than a single food, it is not evidence for fish or for coconut oil on their own.

Fruit and vegetables in total were not related to any semen parameter in the Boston cohort (Chiu 2015). The authors found an association only after splitting produce by its typical pesticide-residue load, which is taken up in the questions below.

Dairy runs the other way, and only for full-fat. Among 189 university men in Rochester, New York (Afeiche 2013), those in the upper half of full-fat dairy intake had 3.2 percentage points fewer normally shaped sperm than the lower half (95% CI -4.5 to -1.8), driven mostly by cheese; the same contrast for low-fat dairy was -1.3 points and not statistically significant. One cross-sectional sample, one semen sample per man.

The Mediterranean pattern as a whole, which is roughly what these associations add up to, was linked in a 2025 meta-analysis of eight observational studies (1,835 men; Agarwal) to a higher sperm count, total motility and progressive motility, with heterogeneity between studies of 86 to 89% across the pooled estimates. The one prospective study in that review that followed couples through assisted reproduction found no association between the man's Mediterranean score and live birth or any semen parameter.

Processed meat appears in both of the Chavarro group's cohorts. Among the 189 Rochester men (Afeiche 2014, Epidemiology), total sperm count fell step by step across quarters of processed red meat intake, with the top quarter 78 million lower than the bottom (95% CI -202 to -5); the authors said they could not separate this from the shorter abstinence times of heavy meat eaters. In the Boston clinic cohort, the top quarter of processed meat intake had 1.7 percentage points fewer normally shaped sperm than the bottom quarter (95% CI -3.3 to -0.04).

Sugar-sweetened drinks: in the same Rochester men (Chiu 2014), those drinking 1.3 or more servings a day had progressive motility 9.8 percentage points lower than men under 0.2 servings a day (95% CI 1.9 to 17.8). The association was present in lean men and absent in overweight men, and no other parameter was affected.

Trans fats: in 209 university students in Murcia, Spain (Chavarro 2014), adjusted mean total sperm count ran from 144 million in the lowest quarter of trans-fat intake to 89 million in the highest (P-trend 0.03); apart from a weak link between cholesterol intake and lower ejaculate volume, nothing else in the diet reached significance, and the authors wrote that they did not know whether the difference in count translates into a difference in fertility.

Alcohol: a 2017 meta-analysis of 15 cross-sectional studies (16,395 men; Ricci) found semen volume 0.25 mL and normal morphology 1.87 percentage points higher in men with no or low intake than in drinkers, with the gap sitting between occasional and daily drinkers rather than between never and occasional. The authors' only advice was to avoid heavy drinking. The alcohol question on its own is covered in does alcohol reduce sperm count.

Cola and other caffeinated soft drinks: a systematic review of 28 observational studies (19,967 men; Ricci 2017) found some studies linking them to lower semen volume, count and concentration, while caffeine from coffee, tea and cocoa was not associated with semen parameters in most studies. The cola signal overlaps with the sugar-sweetened-drink finding above, and neither design can separate the sugar from the caffeine from the rest of a diet that includes a lot of both.

For a large share of Indian readers this is the real question, and the honest answer is that the data are thin and do not agree with each other. A 2024 meta-analysis of observational studies (Samimisedeh, 16 studies pooled) found no significant difference between vegetarian and omnivorous men in total sperm count, concentration, total or progressive motility, or morphology. The single study most often quoted against vegetarian diets, from a clinic in Loma Linda, California (Orzylowska 2016), compared 26 life-long lacto-ovo vegetarians and 5 vegans with 443 non-vegetarians and found lower concentration (50.7 vs 69.6 million/mL) and lower total motility (33.2% vs 58.2%) in the lacto-ovo group, with similar morphology and chromatin integrity; the vegetarian groups were small and the authors said the results should be read cautiously. The one Indian comparison on record is old: in 1976, Jathar and colleagues found that lacto-vegetarian men had lower vitamin B12 in seminal plasma than non-vegetarians but no difference in sperm count, motility or the share of abnormal forms among men with normal counts.

Soy has been studied three ways. In 99 men at the Boston clinic (Chavarro 2008), the highest soy-food intake category had a sperm concentration 41 million/mL lower than men eating no soy (95% CI -74 to -8), with motility, morphology and volume unaffected and the effect concentrated in overweight men. A randomised crossover trial in healthy men (Beaton 2010) fed milk protein, low-isoflavone soy protein and high-isoflavone soy protein for 57 days each and found no change in volume, concentration, count, motility or morphology. And when 184 men going through IVF were followed by soy intake (Mínguez-Alarcón 2015), live birth per cycle did not vary with it. Taken together: one observational signal on concentration, one small trial and one IVF cohort showing nothing. There is no reason in these data to remove soy, tofu or soy milk from a plate, and no evidence that adding them helps.

Not this month. The sperm in a sample given today started forming about two months ago. When 11 men with normal counts drank isotope-labelled water for three weeks and then gave samples every two weeks (Misell 2006), labelled sperm first appeared in the ejaculate after a mean of 64 days (SD 8, range 42 to 76); in every man but one it was at least 60 days. Amann's 2008 review of the older histological data argues that the accepted textbook figure for the whole process might be in error by about six days. Either way, the nut trials ran 12 and 14 weeks for a reason.

So a semen analysis repeated a few weeks after changing what you eat is still measuring the old diet. If you want to know whether a change has done anything, the comparison needs the same laboratory, the same abstinence interval and at least one full cycle of sperm production between samples. How to set that up is in how to prepare for a semen analysis, and how to read the result line by line is in the semen analysis report guide.

Food is a reasonable thing to change while waiting for a repeat test. It is not a treatment for the conditions that most often sit behind an abnormal result. A varicocele, an obstruction, a hormonal or genetic cause, or azoospermia (no sperm in the ejaculate) each has its own pathway, and none of the studies above included men with those diagnoses as a group. The range of causes is set out in what causes low sperm count.

What to do with a result: if the first analysis was abnormal, a repeat after the interval above comes before any conclusion; if the repeat is still abnormal, the next step is an assessment of cause, not a longer diet. Whether the supplements built on these same nutrients work, and whether changing diet before treatment is worth the time, is the subject of a separate page: do foods or supplements actually improve sperm quality.

Being handed a diet sheet after a bad result can feel like being told the problem is something you did. Nothing in the evidence above supports that reading: the effects measured are modest, most are associations, and none of these studies could tell a particular man why his numbers are what they are.

• That any food raises the chance of pregnancy or live birth. The two nut trials did not measure pregnancy; the nut cohort in 229 couples having assisted reproduction (Salas-Huetos 2022, as tabulated by Cardoso 2024) and the Mediterranean-diet cohort of 245 men (in Agarwal 2025) found no association with live birth; soy intake in 184 IVF men was unrelated to live birth.
• That a change in a semen parameter is a change in fertility. Motility, morphology and vitality moved by about half a standard deviation in the nut trials; the trials did not test whether that moved anything else.
• That fish, low-fat dairy, processed meat, sugary drinks, trans fats, alcohol or soy cause the differences seen. Every one of those findings is cross-sectional or cohort data from food-frequency questionnaires, mostly from one Boston clinic or one Rochester student sample, each with its own confounders: abstinence time, body weight, the rest of the diet.
• That the Mediterranean-diet findings are secure. The 2024 systematic review (Piera-Jordan, 10 studies, 2,032 men) graded every included study very low, low or moderate on GRADE, 'as they were largely observational and contained some bias'; the 2025 meta-analysis found 77 to 89% heterogeneity across its pooled estimates.
• That results in healthy students apply to men with an abnormal semen analysis. Both nut trials and the Rochester and Murcia samples recruited healthy young men; the Boston and Vellore cohorts recruited fertility-clinic patients, but only observationally.
• That any Indian spice, herb or traditional preparation changes a semen analysis. None was tested in any study retrieved for this page.